Extensive neurological recovery from a complete spinal cord injury: a case report and hypothesis on the role of cortical plasticity

Abstract

Neurological recovery in patients with severe spinal cord injury (SCI) is extremely rare. We have identified a patient with chronic cervical traumatic SCI, who suffered a complete loss of motor and sensory function below the injury for 6 weeks after the injury, but experienced a progressive neurological recovery that continued for 17 years. The extent of the patient's recovery from the severe trauma-induced paralysis is rare and remarkable. A detailed study of this patient using diffusion tensor imaging (DTI), magnetization transfer imaging (MTI), and resting state fMRI (rs-fMRI) revealed structural and functional changes in the central nervous system that may be associated with the neurological recovery. Sixty-two percent cervical cord white matter atrophy was observed. DTI-derived quantities, more sensitive to axons, demonstrated focal changes, while MTI-derived quantity, more sensitive to myelin, showed a diffuse change. No significant cortical structural changes were observed, while rs-fMRI revealed increased brain functional connectivity between sensorimotor and visual networks. The study provides comprehensive description of the structural and functional changes in the patient using advanced MR imaging technique. This multimodal MR imaging study also shows the potential of rs-fMRI to measure the extent of cortical plasticity.

Keywords: diffusion tensor imaging; magnetization transfer imaging; plasticity; resting state fMRI; spinal cord injury; trauma.

Figure 1

ASIA classification of the patient after recovery, performed according to the International Standards for Neurological Classification.

Figure 2

Structural MR images acquired from the patient (top row) and a control (bottom row). (A,E) Sagittal T2-w images. Bracket in (A) highlights injury epicenter of the patient. (B,F) Axial T2-w images. (C,G) Axial MTCSF images. (D,H) Axial FA color maps. The FA maps are color-coded using a standard diffusion color-encoding scheme (r/l: red, a/p: green, s/i: blue). Each row of (B–D) shows a section from: (1) above the injury epicenter, (2) injury epicenter, and (3) below the injury epicenter. Images from the corresponding cervical levels in the control are shown in each row of (F–H).

Figure 3

Quantification of whole cord and white matter atrophy, as a measure of white matter tissue sparing throughout the cervical spinal cord. (A) Degree of whole cord atrophy, as measured by the whole cord area, (B) white matter area, and (C) the ratio of wma over wca, was measured between C2–C6. Values of the controls (blue line) are mean ± SD. Plot of wca (A) and wma (B) demonstrate atrophy of the patient's spinal cord not only at the injury epicenter (dark area) but throughout the cervical cord. Plot of ratio of wma over wca (C) shows marked decrease of white matter around the injury epicenter.

Figure 4

Column-specific data comparison of DTI- and MTI-derived quantities of the controls (blue line) and the patient (red line). FA, MD, λ_ǁ, and λ_⊥, and MTCSF values of each spinal column were measured between C2–C6. The resulting column profiles are shown. Darker area within the plots indicates the patient's injury epicenter. Values reported for the controls are mean ± SD. FA value decreased, while MD, λ_ǁ, and λ_⊥ increased at the injury epicenter. MTCSF value increased throughout the cervical cord.

Figure 5

Functional networks of the controls and the patient estimated using GICA. GICA was performed to delineate same functional networks in the controls and the patient. ICs are overlaid on the Montreal Neurological Institute (MNI) template. (A–L) First column corresponds to the sagittal, coronal, and axial view of the controls' functional networks. Second column corresponds to the patient's functional networks. Coordinates (in mm) for each view are indicated on top of the subfigures, along with the IC numbers assigned during GICA. WNC of the controls (in blue) and the patient (in red) for each network are shown as a series of box plots. On each box, central mark is the median and edges of the box are the 25th and 75th percentiles. Aud, auditory; Smot, seonsorimotor; Vis, visual; DMN, default mode network; Attn, attention; Exec, executive; Sal, salience.

Figure 6

Between-network connectivity. (A) BNC correlation matrix shows synchrony between pairs of functional networks. (B) Combined matrix of the standard deviation of the corresponding BNC measurements. (A,B) In reference to the top–left to bottom–right diagonal axis, bottom–left portion corresponds to the controls and top–right portion corresponds to the patient. Diagonal elements have been zeroed for display purposes. (C) Difference between the controls and the patient's BNC measurements. (D) Box plots of controls' BNC measurements, with the patient's mean BNC values indicated by red dots. Shown are the 10 functional networks with the largest differences between the patient and controls.